Selanikio flags DTC labs + ChatGPT

Consumer health just took another step away from the doctor’s office. The new stack is pretty clear now — wearable companies sell you data, lab startups sell you bloodwork, and ChatGPT helps turn the results into something that feels legible. That is the setup Joel Selanikio has been pointing at. The news under it is concrete: OpenAI launched ChatGPT Health on January 7, 2026, and Function launched a ChatGPT app the same day, right as direct-to-consumer lab testing was already spreading across Oura, WHOOP, and Hims. (openai.com) ### What changed here? For years, people could buy wearables, Google symptoms, and order some tests online. But those pieces were fragmented. The shift now is that the pieces are starting to connect. ChatGPT Health gives users a dedicated place inside ChatGPT to connect health data and apps, while Function’s app lets members share a high-level summary of lab results into ChatGPT for mo(openai.com)to something closer to a consumer health interface. (openai.com) ### Why is Function at the center? Function is the cleanest example because it sells a broad annual testing membership and is explicitly building around AI. The company says its membership includes 160+ lab tests for $365 a year. In November 2025, it raised $298 million in a Series B at a $2.5 billion valuation. That funding round matters because it says investors think this is not a quirky wellness niche anymore — it is a platform business. (functionhealth.com) ### Are Oura, WHOOP, and Hims doing the same thing? Not exactly, but they rhyme. Oura’s Health Panels let members order testing for 50 biomarkers inside the Oura app. WHOOP Advanced Labs ties 122+ biomarkers to continuous wearable data, with Quest handling the lab side. Hims launched direct-to-consumer testing through Quest too, offering up to 120 biomarker tests across 10 health categories. Differ(functionhealth.com)st, then software interprets the result. (ouraring.com) ### Why does ChatGPT matter so much? Because the bottleneck was never just getting the test. It was making sense of the output. A lab report is basically a spreadsheet with emotional consequences. ChatGPT Health is built to pull together records, app data, and health questions in one place, and OpenAI says more than 230 million people already ask health and wellness questions on ChatGPT each wee(ouraring.com)ng up to it. (openai.com) ### What is Selanikio really flagging? He is not just saying “AI in healthcare.” He is pointing to care unbundling. The old flow was simple: doctor orders test, lab runs test, doctor explains result, chart stores result. The new flow can start with the patient, move through a consumer app, get interpreted by AI, and only later touch a clinician — if it touches one at all. That changes who owns the first conversation about health. (futurehealth.live) ### Where does the EHR problem show up? Right at the handoff. If a patient brings in self-ordered labs, wearable trends, and an AI-generated interpretation, someone has to decide what belongs in the medical record, what can be trusted, and who is responsible for follow-up. OpenAI built ChatGPT Health with extra privacy controls and says those chats are isolated from regular ChatGPT use, but that d(futurehealth.live)ems. (openai.com) ### What is the catch? More data does not automatically mean better care. Direct-to-consumer testing can surface useful signals early, but it can also create false alarms, incidental findings, and a lot of anxious interpretation. That tension is already showing up in mainstream coverage of consumer bloodwork — easy access is here, but the meaning layer is still messy. (apr.org)-work-now-interpreting-the-results-is-another-story)) ### Bottom line? The important change is not any single product. It is that testing, context, and interpretation are now getting stitched together outside the clinic. If that stack keeps improving, primary care and EHR vendors will not just need integrations — they will need a new theory of where care starts.